Automated Zebrafish Chorion Removal and Single Embryo Placement: Technology Brief 432197

advertisement
432197
ndrell et al.Journal of Laboratory Automation
JLAXXX10.1177/2211068211432197Ma
Technology Brief
Automated Zebrafish Chorion
Removal and Single Embryo Placement:
Optimizing Throughput of Zebrafish
Developmental Toxicity Screens
Journal of Laboratory Automation
17(1) 66–74
© 2012 Society for Laboratory
Automation and Screening
DOI: 10.1177/2211068211432197
http://jala.sagepub.com
David Mandrell1, Lisa Truong1, Caleb Jephson1, Mushfiqur R. Sarker1,
Aaron Moore1, Christopher Lang1, Michael T. Simonich1, and Robert L.Tanguay1
Abstract
The potential of the developing zebrafish model for toxicology and drug discovery is limited by inefficient approaches to
manipulating and chemically exposing zebrafish embryos—namely, manual placement of embryos into 96- or 384-well plates
and exposure of embryos while still in the chorion, a barrier of poorly characterized permeability enclosing the developing
embryo. We report the automated dechorionation of 1600 embryos at once at 4 h postfertilization (hpf) and placement
of the dechorionated embryos into 96-well plates for exposure by 6 hpf. The process removed ≥95% of the embryos from
their chorions with 2% embryo mortality by 24 hpf, and 2% of the embryos malformed at 120 hpf. The robotic embryo
placement allocated 6-hpf embryos to 94.7% ± 4.2% of the wells in multiple 96-well trials.The rate of embryo mortality was
2.8% (43 of 1536) from robotic handling, the rate of missed wells was 1.2% (18 of 1536), and the frequency of multipicks
was <0.1%. Embryo malformations observed at 24 hpf occurred nearly twice as frequently from robotic handling (16 of
864; 1.9%) as from manual pipetting (9 of 864; 1%). There was no statistical difference between the success of performing
the embryo placement robotically or manually.
Keywords
zebrafish, automation, chorion, robotic
Introduction
Toxicology is undergoing a paradigm shift recognized by
the U.S. Environmental Protection Agency, the National
Toxicology Program, and the National Research Council.
This is nowhere more evident than the Tox21 agenda1,2 and
its European counterpart, the REACH initiative,3 which will
fundamentally rely on high-throughput screening methods
made possible by enormous advances in robotics, digital
imaging, and computational and informational tools to
assess a staggering backlog of more than 60 000 chemicals
now in production, for most of which little if any toxicology
data exist. Although the Tox21 vision emphasizes complete
transition to in vitro screening for toxicity prediction, for the
foreseeable future, rapid in vitro methods will not eliminate
animal toxicity testing but serve to prioritize chemicals for
further screening in animals. However, a bottleneck of time
and money stands in the way of evaluating even a small
fraction of the in vitro prioritized chemicals with conventional rodent models of human risk. A solution to the
bottleneck is much faster, predictive animal models that are
amenable to automation platforms. Zebrafish has emerged
as the choice for such a model. No other vertebrate is better
suited to high-throughput chemical screening.4
Zebrafish development is the most sensitive life stage
to chemical exposure. The bulk of the model’s utility, and
hence the bulk of zebrafish toxicology work, is centered
on development. Because zebrafish embryos remain transparent throughout much of organogenesis, adverse effects of
chemical exposure on development of the brain, notochord,
heart, jaw, body segmentation, and body size can be continuously assessed in the living animal under low magnification.
An important developmental feature is that zebrafish embryos
that are malformed, missing organs, or displaying organ
1
Oregon State University, Corvallis, OR, USA
Received Sep 26, 2011.
Corresponding Author:
Robert L. Tanguay, Department of Environmental & Molecular Toxicology,
Oregon State University, 28645 East Hwy. 34, Corvallis, OR 97333, USA
Email: Robert.Tanguay@oregonstate.edu
Downloaded from jla.sagepub.com by guest on January 28, 2012
67
Mandrell et al.
dysfunction usually survive well beyond the point at which
those organs normally start to function. This feature is in
stark contrast to rodents, in which heart and other organ malformations, as well as missing or dysfunctional organs, typically cause a generalized in utero lethality. In a large-scale
(>1000 animal) rodent screen, such end points would be
missed with anything less than Herculean efforts at
detection.4
As the only vertebrate model meeting the rapid, predictive toxicology needs of the 21st century, zebrafish are
increasingly used by public- and private-sector interests to
conduct discovery screens of chemical libraries containing
≥1000 compounds (reviewed in Sukardi et al.5 and Zon and
Peterson6; see also Kokel et al.,7 Paik et al.,8 Rihel et al.,9
Wang et al.,10 and Yen et al.11). These same studies also highlight near-exclusive reliance on inefficient approaches to
manipulating and chemically exposing zebrafish embryos—
namely, manual placement of embryos into 96- or 384-well
plates and exposure of embryos while still in the chorion, an
acellular barrier of poorly characterized permeability enclosing the developing embryo.
Manual placement of embryos to microtiter wells is not
a cost-effective use of laboratory personnel and, although
barely feasible for screens of a few thousand compounds, is
completely impractical for the scale needed to address the
rapidly growing backlog of conventional and nanomaterial
chemistries already in use. When working with dechorionated embryos, manual placement by humans requires a considerably more refined handling technique than embryos in
the chorion. High-precision, repetitive motions necessitate
many breaks for lab technicians, and the mundane nature of
the task equates to frequent personnel turnover and inefficiencies associated with continual retraining.
No chemically comprehensive assessment of chorion
permeability has been reported, but the chorion is widely
suspected to influence chemical uptake, and several reports
confirm that it is an uptake barrier for metal nanomaterials.12,13 Reported attempts at zebrafish chorion removal prior
to 24 h postfertilization (hpf) on a large (>100 embryo) scale
have been plagued by generally low survival,14 and exposure
in the chorion continues to be a common practice in large
screens. Such screens have yielded much information and
potentially invaluable therapeutic discovery (reviewed in
Zon and Peterson6), but it is tempting to speculate how much
information has been missed because of permeability limitations of the chorion. A second, important limitation of not
removing the chorion prior to exposure is that compounds
that specifically inhibit the hatching process lead to secondary phenotypic responses. For example, the widely used
insecticide cartap inhibits hatching, resulting in secondary
effects on the embryo such as wavy notochord, axis malformation, and somite defects.15 These effects are due to the lack
of hatching, rather than a primary response to exposure to the
chemical. The chorion can significantly confound the early life
stage zebrafish toxicity assay by leading to false positives.
When we consider the obvious potential for false negatives
due to the aforementioned barrier effect, chorion removal is
critical to improve the predictivity of the assay.
Herein we report rapid and cost-effective automated
removal of the chorion from 2000 embryos at once at the
4-hpf stage and placement of the dechorionated embryos
into 96-well plates for exposure at 6 hpf. Two approaches
were key to developing these platforms: (1) the use of pronase degradation of the chorion combined with automated
agitation and washing of the embryos and (2) the application of machine vision-guided robotics to rapidly select and
place the dechorionated embryos into plate wells. A single
station is used in our laboratory to plate >1000 dechorionated embryos per day, requiring approximately 4 h and with
a survival rate better than 95% by 120 hpf.
Materials and Methods
Zebrafish
Embryonic zebrafish were obtained from a Tropical 5D strain
of zebrafish (Danio rerio) reared in the Sinnhuber Aquatic
Research Laboratory (SARL) at Oregon State University,
Corvallis. Adults were kept at standard laboratory conditions
of 28 °C on a 14-h light/10-h dark photoperiod in fish water
(FW) consisting of reverse-osmosis water supplemented with
a commercially available salt (Instant Ocean, www.instant
ocean.com) to create a salinity of 600 microsiemens. Sodium
bicarbonate was added as needed to adjust the pH to 7.4.
Zebrafish were group spawned, and embryos were collected
and staged as described by Kimmel et al.16
Automated Chorion Removal
To establish an inexpensive and highly reproducible method
of removing chorions from about 1600 embryos at a time at
4 hpf, a Belly Dancer shaker (ATR, Inc., Laurel, MD) was
modified to accommodate a custom-machined, anodized
aluminum shaker plate that holds 4 glass Petri dish bottoms
(100 × 15 mm; VWR, Radnor, PA) and attached water
delivery tubing, stainless steel nozzles, and a drain port
(Fig. 1). The internal workings of the Belly Dancer were
modified with a small pump and a parametric motion controller (Revolution Robotics, Inc., Corvallis, OR). The front
panel was modified with an LED display and push button
control. The on-board pump supplied rinse water from an
external heated (28 °C) carboy via the tubing and nozzles
to each glass dish at the appropriate time. The movement of
the shaker was controlled by the same system to deliver
pulsed agitation or gentle swirling, precisely when needed,
to dislodge partially hydrolyzed chorions. The only manu-
Downloaded from jla.sagepub.com by guest on January 28, 2012
68
Journal of Laboratory Automation 17(1)
10-min process for a trained technician. Approximately 400
to 500 embryos were placed in each of the 4 glass dishes in
25 mL of FW with 50 µL of 50 mg/mL pronase (#81748;
Fluka, St. Louis, MO) for 6.5 min while the dechorionator
platform constantly agitated. The pronase was then flushed
away by gently overflowing the dish with the pumped-in
FW for 10 min with 45-s agitation cycles separated by 15 s
while still. The total volume of fish water consumed was
about 1 L. After the pronase and rinse phases, the embryos
were incubated for 20 min at 28 °C, agitated once more
to dislodge any remaining chorions, and rinsed again to
remove the dislodged chorions. The dechorionation was
evaluated (Table 1) by gently removing approximately
100 embryos with a flame-polished Pasteur pipette after the
final rinse and examination under a dissecting microscope
for pronase or mechanical damage. No further cleanup of
the dechorionated batches was performed prior to allocation to 96-well plates.
Automated Allocation of Dechorionated
Embryos to 96-Well Plates
Figure 1. A modified shaker platform-based instrument for the
automation of chorion removal from zebrafish embryos at 4 h
postfertilization. (A) A frontal view of the modified Belly Dancer
shaker. The custom machined and anodized aluminum plate is at
top and holds 4 × 100-mm glass dishes. The control panel consists
solely of a start-and-stop button and a small LCD status display.
(B) A closer view of the shaker platform during a rinse phase
of chorion removal. Embryos are visible in the plates. The rinse
water is pumped via the onboard pump from an offboard heated
source and delivered via the hose and nozzle assembly to gently
overflow the plates and not suspend the embryos. The rinse water
is channeled to a drain port at the rear of the platform. Agitation
and pump control are via the custom onboard microcomputer.
ally performed steps were the addition of a pronase aliquot
to commence digestion and pressing of the start button.
No other steps were necessary to operate the device. The
pronase digestion of the chorion was performed at 4 hpf.
Approximately 2000 zebrafish Tropical 5D strain embryos
were received from the Sinnhuber Aquatic Research
Laboratory’s mass spawning facility in a 135-mm plastic
dish and quickly cleaned by removing all dead, unfertilized,
or obviously abnormal embryos with an aspirator, a 5- to
After the rest period, dechorionated embryos at approximately 5 to 6 hpf were transferred to individual wells of a
96-well BD Falcon (BD Biosciences, Franklin Lakes, NJ)
tissue culture polystyrene plate by a custom robotic pick and
place system (Fig. 2; video of the robotic system in operation can be viewed at http://tanguaylab.com/Automation.
html). We noted that the use of nontissue culture-treated
polystyrene plates caused rapid disintegration of 100% of
embryos once removed from the chorion. The system consisted of a four-axis Selective Compliant Assembly Robot
Arm (SCARA; Denso, Inc., Long Beach, CA) with a custom
end-effector designed to replicate a handheld, wide-bore,
flame-polished Pasteur pipette (Fig. 2). A 100-mm glass
Petri dish with approximately 400 embryos was loaded into
a well-lit area below a rigidly mounted machine vision
camera (Allied Vision, Inc., Stadtroda, Germany). Custom
software was used to determine the precise coordinates of a
suitable embryo, which were then passed to the robot control
unit. Under the lighting conditions used, normal embryos
appeared semitransparent, whereas dead embryos appeared
bright white, a parameter easily distinguished by the machine
vision. The robot was programmed to first draw 100 µL of
embryo medium into the flame-polished pipette from a filling station at the beginning of each cycle, drive the pipette to
coordinates several centimeters above the machine visionselected embryo, and place the pipette tip 20 µm above the
embryo coordinates. The embryo was gently aspirated along
with an additional ~20 µL of embryo medium into the pipette
via the onboard syringe pump, and the robot returned the
pipette tip to the coordinates of the liquid surface of the next
empty well of the plate. The wells had been prefilled with
Downloaded from jla.sagepub.com by guest on January 28, 2012
69
Mandrell et al.
Table 1. Results of Automated Chorion Removal at 4 h Postfertilization (hpf), observed at 24 and 120 hpf
Results
Trial Date Embryos Sampled from ≈1600a Normal Dead at 24 hpf Malformed at 24 hpf Dead at 120 hpfb Malformed at 120 hpfc
7/18/2011
7/25/2011
7/26/2011
7/27/2011
7/28/2011
7/29/2011
Total
%
86
100
100
100
97
99
582
100.0
83
97
98
92
95
95
560
96.2
3
3
1
1
2
0
10
1.7
0
0
1
7
0
4
12
2.1
0
0
1
6
1
3
11
1.9
1
1
2
2
2
1
9
1.5
a
The dechorionator holds 4 × 100-mm dishes with approximately 400 embryos each, from which roughly 100 were removed at once by Pasteur pipette
after the process for evaluation.
b
Additional larval mortality at 120 hpf. Excludes mortality observed at 24 hpf.
c
Additional larval malformation at 120 hpf. Excludes malformation observed at 24 hpf.
50 µL of embryo medium. A quick, gentle touch of the liquid surface was all that was required to cause the embryo to
be released to the well by capillary action. Positive dispensing pressure was not needed as the embryo had settled to the
bottom of the liquid column while the robot pivoted between
the source plate and 96-well plate. We noted that an additional 2 to 3 µL of embryo medium was transferred to the
well by capillary release of the embryo. The cycle was completed with an aspirate and total dispense step at a wash
station and a subsequent 100-µL recharge at the filling station. The cycle was then repeated 95 times. Because of the
affinity of dechorionated 6-hpf embryos for each other, the
source Petri dish had to be given a brief, 1-s swirl once during the loading of a plate to redisperse the embryos, improving machine vision selection.
Statistical Analysis
Performance of the robotic embryo placement was evaluated over 16 trials in parallel with manual embryo loading
of 96-well plates. The frequencies of successful well allocations between each method were compared by one-way
analysis of variance (ANOVA) where p < 0.05 was the
threshold for no significant difference between the two
methods.
Results
Automated Chorion Removal at 4 hpf
The performance of six trials of enzymatic chorion removal
from approximately 1600 zebrafish embryos at once was
assessed from random samples of about 100 embryos from
each trial and summarized in Table 1. In each trial, at least
95% of the 4-hpf embryos were successfully removed from
the chorion where success was defined as alive at 24 hpf
with no malformations evident by 120 hpf. The automated
chorion removal resulted in only about 2% embryo mortality by 24 hpf, and only 2% of the embryos were malformed
at 120 hpf. Figure 3 depicts the type of mortal damage that
1%–2% of the embryos were observed to have immediately
following the automated dechorionation process. Early
damage that may have led to 2% of the embryos being
malformed at 120 hpf was not visibly detected.
Automated Placement of Dechorionated
Embryos into 96-Well Plates
We evaluated the success of robotic 96-well loading of
embryos dechorionated at 4 hpf that were 5–6 hpf at the
time of plate loading. The standard for comparison was
our routine method of manual placement of dechorionated
embryos using a handheld Pasteur pipette. Sixteen 96-well
plate comparisons were performed in parallel, and the
results are summarized in Table 2. The manual loading data
were derived from several of the personnel in our laboratory who are equally adept at the technique. The robotic
system successfully allocated embryos to 94.7% ± 4.2% of
the wells, and manual loading successfully allocated embryos
to 94.9% ± 3.6% of the wells. There was no statistical difference between the success of the two methods (ANOVA F =
0.53; p < 0.01 that a significant difference existed). The criterion for success was that each well received only 1 embryo
and that the embryo was alive and not visibly damaged or
malformed at 24 hpf. We note that 2.8% (43 of 1536) of
the unsuccessful wells were from mortality directly as a
result of robotic handling. Mortality from manual loading
accounted for 4% (62 of 1536) of the unsuccessful wells.
The robotic system missed 1.2% (18 of 1536) of the
wells, but only one miss occurred from manual loading.
The frequency of multipicks (2 embryos allocated to a
single well) was similar for the robotic and manual loading
Downloaded from jla.sagepub.com by guest on January 28, 2012
70
Journal of Laboratory Automation 17(1)
Figure 2. A custom robotic station for the automated transfer of dechorionated zebrafish embryos from the dish in which chorion
removal occurred to a 96-well plate. (A, B) A front and side view of the workstation. An extruded aluminum strut assembly rigidly
supports the overhead camera. The station is supported from below by a steel tooling plate welded to a rigid steel table. The four-axis
Denso SCARA robot is bolted to the tooling plate as it generates strong inertial forces during its movement.The syringe pump supplying
the pipetting force is visible next to the base of the robot. Also visible are the lighted arena beneath the camera and the holder for the
destination plate and a second plate for pipette rinsing and preloading with water. (C) The custom end effecter gently cradles a single,
wide-bore, flame-polished Pasteur pipette connected to the syringe pump line. (D) The lighted arena consists of an aluminum square
with precisely located pins in the corners that serve as visible references for the machine vision software.The circular array of LED units
provides the optimal amount of contrast needed for the machine vision to clearly see the embryos. The units are commercially available,
high-intensity LED assemblies. When operating, the station is protected by a light curtain that, if interrupted, stops the robot’s motion in
less than 50 ms.
Downloaded from jla.sagepub.com by guest on January 28, 2012
71
Mandrell et al.
Figure 3. Mortal damage typically observed in 1%–2% of 5–h
postfertilization (hpf) zebrafish embryos immediately after the
automated dechorionation process. The damaged embryos are
indicated by arrows; all other embryos in the field are normal.
Whether the low-frequency damage is due to the effects of
pronase digestion or motion is unknown.
(3 and 2 embryos, respectively, out of 1536). Embryo malformations observed 18 h after plate loading (24 hpf)
occurred nearly twice as frequently from robotic handling
(16 of 864; 1.9%) as they did from manual pipetting (9 of
864; 1%). No bias toward dead or malformed embryos
occurring in certain wells was ever observed.
Discussion
We have introduced automated platforms for highthroughput chorion removal at 4 hpf and 96-well plate
allocation at 6 hpf that consistently yielded 95% healthy
embryos. Together, these automation platforms provide (1)
a rapid and inexpensive circumventing of the potential for
false-negative and false-positive results imposed by the
chorion on high-throughput applications of the developmental zebrafish model and (2) a much less labor-intensive
and more reliable means of carefully allocating dechorionated embryos to 96-well plates at a rate amenable to highthroughput screening.
This is the first report of an en masse chorion removal
method for zebrafish embryos prior to 24 hpf, with a reproducibly high survival rate. Although chorion removal at 24
hpf from 50 to 100 zebrafish embryos at once has been
reported, initiating embryo exposure so late in development
is likely to be of limited utility for large-scale screens. Such
screens lack a priori knowledge of compound activity and
must therefore maximize opportunities for chemical “hits”
by chemically exposing during the widest practical window
of development. A recent report sought to quantify the success of pronase-supported dechorionation at 6 hpf, as
described by Westerfield,17 for replicates of 50 embryos.14
That study concluded that the use of pronase was generally
damaging to 6-hpf embryos and demonstrated a normal
development rate of only 75% and a mortality rate of nearly
40%.14 Our demonstration of pronase-supported, automated
chorion removal, at 4 hpf, from 1600 embryos at once, consistently yielded ≥95% survival and normal development to
120 hpf, indicating that a pronase-supported approach can
be both practical and scalable to meet the embryo demands
of high-throughput screening.
To our knowledge, this is also the first report of a reproducible method for robot-automated allocation of embryos
to microtiter plates. A recent report described an imagebased fluidic approach to rapid allocation of embryos to
96-well plates.18 Evidence of that system’s performance
was largely limited to the handling of embryos still in the
chorion with only cursory performance data from dechorionated embryos. We have demonstrated the highly reproducible use of a small, industrial robotic arm approach to
retrieve single dechorionated embryos at 6 to 7 hpf from an
unsorted dish and allocate them to a 96-well plate with a
better than 95% survival rate. It would also be straightforward and, in some instances, desirable to allocate more than
one embryo per well, such as for monitoring subtle locomotor activity where a higher signal-to-noise ratio is achieved
with multiple embryos.7 For gross malformation end points,
one embryo per well minimizes the potential effects of dose
titration from uptake by multiple animals in the same 100µL volume. Another practical extension of the automation
would be for sorting transient transgenic reporter animals
fluorescing at the embryonic stage and fluorescent-tagged
morpholino-injected embryos in high-throughput gene
knockdown assays.
We did not include the time to complete plate loading for
any of the robotic trials shown in Table 2, focusing instead
on the ability of the modified SCARA robot to handle
embryos gently with high survival and low malformation
rates. Figure 4 summarizes the entire process with approximate times for completion. We note that although several
variables affected the time required for the robotic system
to complete a 96-well plate, by far the most important variable was density and dispersal of embryos in the source
dish. Having more than 300 embryos in the source dish, or
failing to keep the embryos well dispersed with periodic
swirling, noticeably slowed the rate at which the machine
vision camera and software could select a sufficiently isolated embryo to map and direct the robot to retrieve. Once
the source dish was depleted to less than 300 embryos and
the dispersal was kept at a maximum, the system consistently loaded one 96-well plate every 15 min. For comparison, laboratory personnel that perform the task on a daily
basis consistently loaded one 96-well plate every 6 to 10
min. However, three experienced loaders in our laboratory
could only load a total of 18 to 20 plates before fatigue
Downloaded from jla.sagepub.com by guest on January 28, 2012
72
Journal of Laboratory Automation 17(1)
Table 2. Robotic Pick-and-Place (P&P) Performance on 6–h Postfertilization (hpf) Dechorionated Zebrafish Embryos Compared to
Manual Plate Loading
Overall P&P Performance
Trial Date
Wells
8/1/2011
8/2/2011
8/2/2011
8/2/2011
8/4/2011
8/4/2011
8/5/2011
8/8/2011
8/9/2011
9/6/2011
9/6/2011
9/6/2011
9/7/2011
9/7/2011
9/7/2011
9/7/2011
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
Specific P&P Failures
Wells Allocated
Successfully, No. (%)a
Wells Unsuccessful,
No. (%)b
P&P
Mortalityc
Missed
Wellsd
Multipickse
Malformedf
86 (89.6)
90 (93.8)
93 (96.9)
95 (99.0)
85 (88.5)
91 (94.8)
88 (91.7)
87 (90.6)
89 (92.7)
95 (99.0)
91 (94.8)
96 (100.0)
84 (87.5)
95 (99.0)
95 (99.0)
95 (99.0)
Ave = 94.7% ± 4.2%g
n = 16 trials
10 (10.4)
6 (6.3)
3 (3.1)
1 (1.0)
11 (11.5)
5 (5.2)
8 (8.3)
9 (9.4)
7 (7.3)
1 (1.0)
5 (5.2)
0
12 (12.5)
1 (1.0)
1 (1.0)
1 (1.0)
2
4
2
0
9
2
6
6
4
0
0
0
8
0
0
0
2.8%
5
0
0
0
0
1
0
0
2
1
5
0
1
1
1
1
1.2%
0
0
0
0
0
0
0
1
0
0
1
0
3
0
0
0
0.3%
3
2
1
1
2
2
2
2
1
—
—
—
—
—
—
—
1.9%
Overall Manual Performance
Trial Date
h
8/1/2011
8/2/2011h
8/2/2011h
8/2/2011h
8/4/2011h
8/4/2011
8/5/2011
8/8/2011
8/9/2011
9/6/2011
9/6/2011
9/6/2011
9/7/2011
9/7/2011
9/7/2011
9/7/2011
Wells
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
96
Wells Allocated
Successfully, No. (%)
Wells Unsuccessful,
No. (%)
87 (90.6)
89 (92.7)
87 (90.6)
91 (94.8)
86 (89.6)
91 (94.8)
91 (94.8)
88 (91.7)
96 (100.0)
96 (100.0)
96 (100.0)
89 (92.7)
89 (92.7)
96 (100.0)
94 (97.9)
95 (99.0)
Ave = 94.9% ± 3.6%
n = 16 trials
9 (9.4)
7 (7.3)
9 (9.4)
5 (5.2)
10 (10.4)
5 (5.2)
5 (5.2)
8 (8.3)
0
0
0
7 (7.3)
7 (7.3)
0
2 (2.1)
1 (1.0)
Specific Manual Failures
Manual Load
Mortality
Missed
Wells
6
7
8
1
10
5
3
6
0
0
0
7
7
0
1
1
4.0%
a
Well allocation is placement of one dechorionated 6-hpf embryo.
An unsuccessful well allocation (errorsc-e) was determined immediately after the trial and at 24 hpf for.
c
Embryo dead, usually partly or completely disintegrated, immediately after trial.
d
Embryo or disintegrated residue absent from well immediately after trial.
e
More than one embryo allocated to the same well.
f
Malformations observed at 24 hpf; Dash indicates not recorded 9/6–9/7.
g
Robotic performance was not different from the overall success of manual loading (F = 0.53; p < 0.01).
h
Lower apparent performance success of earlier trials reflects the performance of earlier software versions.
b
Downloaded from jla.sagepub.com by guest on January 28, 2012
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
—
Multipicks
1
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0.2%
Malformed
1
0
1
3
0
0
2
2
0
—
—
—
—
—
—
—
1.0%
73
Mandrell et al.
of chemical exposures, and they relieve a serious bottleneck
to high-throughput use of the developing zebrafish by automating embryo allocations to assay plates. We believe that
these advances, coupled with advances in automated imaging and phenotype analysis, will quickly enable researchers
to expand the scale and scope of toxicology and discovery
research.
Acknowledgments
The authors wish to express their gratitude to Greg Gonnerman,
Lindsey Chalker, Carrie Barton, Cari Buchner and Chapell Miller
of the Tanguay Laboratory for providing a steady supply of high
quality zebrafish eggs and for their insights on how to improve the
automation. We thank Dr. Siba Das for Photoshop help with the
figures.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
Funding
The authors disclosed receipt of the following financial support
for the research, authorship, and/or publication of this article: The
presented work was funded in part by NIH grants RC4 ES019764,
P42 ES016465, and P30 ES000210.
References
Figure 4. Process summary and timeline for automated chorion
removal and embryo allocation to plate wells. Step 1, beginning at
4 h postfertilization (hpf), consists of removal of dead or obviously
abnormal embryos from the mass of 2000 embryos in a 135-mm
Petri dish. It requires 5 to 10 min to complete the cleanup using a
Pasteur pipette connected to a vacuum aspirator. Step 2, beginning
immediately after 1, requires the approximate division of the mass
of embryos among 4, 100-mm glass Petri dishes, the addition
of a pronase aliquot to each, and pressing the start button. The
automated gentle shaking, rinsing, and rest period requires 40 to 45
min. Step 3, beginning immediately after the post-dechorionation
rest period and rinse to remove any remaining traces of chorion,
simply requires the movement of the plate of 400 embryos to the
Selective Compliant Assembly Robot Arm (SCARA) robot station
and pressing the start button. A single embryo is delivered to each
of the 96 wells in 15 min.
resulted in a successful loading rate of <95% when the plates
were observed at 24 hpf. The robotic system, facing no such
limitation, offers an obvious advantage when throughput
demands require ≥200 plates per week. Moreover, having
developed the software for a single system, the cost of scaling the system to multiple SCARA robotic loading stations
will be limited to hardware only.
The automation platforms herein obviate persistent concerns about the chorion and its potential to limit the effects
1. Hartung, T. Lessons Learned from Alternative Methods and
Their Validation for a New Toxicology in the 21st Century. J.
Toxicol. Environ. Health B Crit. Rev. 2010, 13(2–4), 277–290.
2. Schmidt, C. W. TOX 21: New Dimensions of Toxicity Testing.
Environ. Health Perspect. 2009, 117(8), A348–A353.
3. Lilienblum, W.; Dekant, W.; Foth, H.; Gebel, T.; Hengstler, J. G.;
Kahl, R.; Kramer, P. J.; Schweinfurth, H.; Wollin, K. M. Alternative Methods to Safety Studies in Experimental Animals:
Role in the Risk Assessment of Chemicals under the New
European Chemicals Legislation (REACH). Arch. Toxicol.
2008, 82(4), 211–236.
4. Zon, L. I.; Peterson, R. T. In Vivo Drug Discovery in the
Zebrafish. Nat. Rev. Drug Discov. 2005, 4(1), 35–44.
5. Sukardi, H.; Chng, H. T.; Chan, E. C.; Gong, Z.; Lam, S. H.
Zebrafish for Drug Toxicity Screening: Bridging the In Vitro
Cell-Based Models and In Vivo Mammalian Models. Expert
Opin. Drug Metab. Toxicol. 2011, 7(5), 579–589.
6. Zon, L. I.; Peterson, R. The New Age of Chemical Screening
in Zebrafish. Zebrafish 2010, 7(1), 1.
7. Kokel, D.; Bryan, J.; Laggner, C.; White, R.; Cheung, C. Y.;
Mateus, R.; Healey, D.; Kim, S.; Werdich, A. A.; Haggarty, S. J.;
et al. Rapid Behavior-Based Identification of Neuroactive
Small Molecules in the Zebrafish. Nat. Chem. Biol. 2010,
6(3), 231–237.
8. Paik, E. J.; de Jong, J. L.; Pugach, E.; Opara, P.; Zon, L. I. A
Chemical Genetic Screen in Zebrafish for Pathways Interacting
Downloaded from jla.sagepub.com by guest on January 28, 2012
74
9.
10.
11.
12.
13.
Journal of Laboratory Automation 17(1)
with cdx4 in Primitive Hematopoiesis. Zebrafish 2010, 7(1),
61–68.
Rihel, J.; Prober, D. A.; Arvanites, A.; Lam, K.; Zimmerman, S.;
Jang, S.; Haggarty, S. J.; Kokel, D.; Rubin, L. L.; Peterson, R. T.;
et al. Zebrafish Behavioral Profiling Links Drugs to Biological
Targets and Rest/Wake Regulation. Science 2010, 327(5963),
348–351.
Wang, C.; Tao, W.; Wang, Y.; Bikow, J.; Lu, B.; Keating, A.;
Verma, S.; Parker, T. G.; Han, R.; Wen, X. Y. Rosuvastatin,
Identified from a Zebrafish Chemical Genetic Screen for Antiangiogenic Compounds, Suppresses the Growth of Prostate
Cancer. Eur. Urol. 2010, 58(3), 418–426.
Yen, J.; Donerly, S.; Levin, E. D.; Linney, E. A. Differential Acetyl
cholinesterase Inhibition of Chlorpyrifos, Diazinon and Parathion
in Larval Zebrafish. Neurotoxicol. Teratol. 2011, 33(6), 735–741.
Johnson, A.; Carew, E.; Sloman, K. A. The Effects of Copper
on the Morphological and Functional Development of Zebrafish
Embryos. Aquat. Toxicol. 2007, 84(4), 431–438.
Bar-Ilan, O.; Louis, K. M.; Yang, S. P.; Pedersen, J. A.;
Hamers, R. J.; Peterson, R. E.; Heideman, W. Titanium
14.
15.
16.
17.
18.
Dioxide Nanoparticles Produce Phototoxicity in the Developing Zebrafish. Nanotoxicology 2011 Aug 10. [Epub ahead
of print]
Henn, K.; Braunbeck, T. Dechorionation as a Tool to Improve
the Fish Embryo Toxicity Test (FET) with the Zebrafish
(Danio rerio). Comp. Biochem. Physiol. C Toxicol. Pharmacol. 2011, 153(1), 91–98.
Zhou, S.; Dong, Q.; Li, S.; Guo, J.; Wang, X.; Zhu, G. Developmental Toxicity of Cartap on Zebrafish Embryos. Aquat.
Toxicol. 2009, 95(4), 339–346.
Kimmel, C. B.; Ballard, W. W.; Kimmel, S. R.; Ullmann, B.;
Schilling, T. F. Stages of Embryonic Development of the
Zebrafish. Dev. Dyn. 1995, 203(3), 253–310.
Westerfield, M. The Zebrafish Book: A Guide for the Laboratory Use of Zebrafish (Danio rerio). 5th ed. University of
Oregon Press: Eugene, 2007.
Graf, S. F.; Hotzel, S.; Liebel, U.; Stemmer, A.; Knapp, H. F.
Image-Based Fluidic Sorting System for Automated Zebrafish
Egg Sorting into Multiwell Plates. J. Lab. Autom. 2011, 16(2),
105–111.
Downloaded from jla.sagepub.com by guest on January 28, 2012
Download